Pith. sign in

REVIEW 4 major objections 5 minor 44 references

Motion Capture from Pan-Tilt Cameras with Unknown Orientation

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Joint optimization of athlete pose and pan-tilt camera rotation, guided by background feature motion, recovers accurate global 3D poses when camera orientations are unknown.

desk verdict A practical sports-mocap pipeline that couples DCT bundle adjustment with ORB homography rotation priors; the central claim holds in the tested setting, but the uncalibrated branch leans on a cue the paper never stress-tests. read the letter →

arxiv 1908.11676 v1 pith:JX33PGTT submitted 2019-08-30 cs.CV

Roman Bachmann , Jörg Spörri , Pascal Fua , Helge Rhodin This is my paper · ORCID
classification cs.CV
keywords markerlessmotioncapturepan-tiltcamerasunknowncameraorientationbundleadjustmenthomography3Dhumanposeestimationalpineskiingbackgroundfeaturetracking
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Alpine ski coaches need the skier's global position, speed, and joint angles, but body-worn sensors are cumbersome and manual annotation is slow, while monocular pose methods cannot recover global scale or position. This paper tries to establish that a multiview markerless system can deliver those global measurements from plain video even when the cameras pan and tilt with unknown orientation. The core proposal is a bundle adjustment that optimizes the 3D pose and the per-frame camera rotations together, while adding a term that ties the rotations to relative motion measured from tracked background features. When camera orientations are estimated this way, local pose accuracy stays close to the fully calibrated case, while the global position error is about 0.7 meters over a large capture volume. The payoff is a practical performance-analysis tool for outdoor sports where fixed-camera calibration is infeasible.

What carries the argument

The central object is the per-frame camera rotation extracted from background homographies, where a homography is the planar projective mapping between two views. For a pan-tilt camera with fixed position, points in consecutive frames $f$ and $f+1$ are related by $H^{f,c} = K^{f+1,c}\,\Delta R^{f,c}\,(K^{f,c})^{-1}$, so the relative rotation $\Delta R^{f,c}$ can be read off from a homography computed from tracked features outside the athlete's bounding box. These measured rotations enter the bundle adjustment through a penalty term that compares them with the rotations produced by the jointly optimized pose, weighted by $\lambda_{\text{rot}} = 10000$. A complementary piece is the discrete cosine basis parametrization of both the pose and the camera angles, which keeps the trajectories smooth and reduces the number of free parameters in the joint optimization.

What would settle it

Run the pipeline on a pan-tilt sequence whose background is textureless, such as a uniform snow slope or blank wall, or has moving people behind the athlete, keep the same optimizer and weights, and compare the global MPJPE against the calibrated result; if the rotation penalty no longer keeps the joint optimization near the calibrated pose, the conditioning claim fails. A simpler version: remove the background-feature rotation term and confirm that the global error jumps toward the reported 7.59 m baseline.

Watch

Extended reading notes

Core claim

The paper's central claim is that the orientation of pan-tilt cameras can be recovered jointly with an athlete's global 3D pose, provided the optimization is conditioned on background feature motion. For a camera fixed in space, consecutive frames are related by a homography that encodes the inter-frame rotation; the paper detects features in the static background outside the athlete's bounding box, computes per-frame rotations from those features, and adds a penalty that keeps the optimized camera rotations consistent with them. With this constraint, the bundle adjustment no longer diverges, and the paper reports, with six cameras and unknown rotations, a global mean per-joint error of $0.701 \pm 0.219$ m, a centered error of $0.090 \pm 0.085$ m, and biomechanical angles within a few degrees of the calibrated result. The same pipeline without the rotation constraint yields a global error of $7.59$ m, and the paper presents the method as the piece that makes uncalibrated multi-camera motion capture practical for fast outdoor sports.

Load-bearing premise

The load-bearing premise is that the background outside the athlete's bounding box is static and textured enough for tracked features to yield accurate per-frame camera rotations, with camera positions and intrinsic parameters known in advance.

Editorial extensions

If this is right

  • Coaches can obtain global center-of-mass position, speed, and joint angles from multi-view video of a fast outdoor sport without camera-orientation calibration, using only relative camera positions and once-calibrated intrinsics.
  • The gap between calibrated and uncalibrated performance is concentrated in global position: centered joint errors and biomechanical angles stay close to the calibrated numbers, so the method is usable for technique analysis even where absolute position matters less.
  • Larger camera counts help the uncalibrated case most on global metrics, with the biggest improvement coming when a third camera is added.
  • Monocular methods that recover only relative pose cannot supply global speed; this approach does, with a reported speed error of about 1.9 m/s against skiers moving at 15–20 m/s.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the conditioning on background homographies is the real cause of the convergence gain, the same mechanism should transfer to other outdoor sports with static scenery, such as cycling or running courses, where fixed camera calibration is equally impractical.
  • The rotation penalty's very large weight suggests that homography noise will translate directly into global drift; a natural extension would be to fuse the same background-feature rotations with cheap inertial angular measurements from the camera to correct low-frequency drift.
  • A testable extension would replace the hand-crafted background features with a learned dense feature matcher, which could extend the method to low-texture or heavily blurred backgrounds, the regime where the paper's own structure-from-motion baseline fails.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes a multi-camera motion capture system for alpine skiing that works with pan-tilt cameras whose orientation is unknown. The pipeline first detects a tight bounding box around the athlete and estimates 2D joint positions with a fine-tuned OpenPose network, using a newly introduced SV-Ski dataset for training. The 3D pose is then recovered by a bundle adjustment that parameterizes both the pose and the camera rotations in a discrete-cosine basis. For the uncalibrated case, the camera rotations are initialized with a look-at scheme and then jointly optimized with the pose under a consistency term that ties inter-frame rotations to homographies computed from ORB features in the static background. On the MV-Ski dataset, the method reports a global MPJPE of 0.092 m in the calibrated case and 0.701 m in the uncalibrated case, with centered errors of 0.077 m and 0.090 m respectively; an ablation without the homography consistency term (Ours-B) degrades to a global MPJPE of 7.59 m. The paper also contributes a new single-view alpine skiing pose dataset.

Significance. If fully validated, the paper addresses a practical and important problem: obtaining global 3D pose and speed measurements for athletes from multiple rotating cameras without cumbersome orientation calibration. The core idea of conditioning camera-rotation optimization on homographies derived from static background features is a sensible alternative to full structure-from-motion, and the DCT parametrization of pose and rotation is a clean way to enforce smoothness. The new SV-Ski dataset is a useful contribution to the community. However, the central claim is currently overstated relative to the actual input requirements, the robust-norm formulation contains a mathematical error, and the main comparison baseline is unpublished. These issues need to be resolved before the performance figures can be taken as fully supporting the stated claims.

major comments (4)
  1. [§4.4, Eq. (8)] The robust reweighting function g(x,y) = (n(0) - n(e_rep)) e_rep is not robust to outliers. Since n(e) tends to 0 as e grows, the prefactor (n(0) - n(e)) saturates to the constant n(0), so the whole expression grows linearly with the residual e_rep rather than downweighting large residuals. The sentence 'Using this norm with σ²=100, outlier points have negligible influence on the energy function' is therefore incorrect. Please replace Eq. (8) with an actual robust norm (for example, a truncated quadratic or a Geman-McClure penalty) or remove the robustness claim; as written, the optimization's tolerance to 2D detection errors is not supported.
  2. [Abstract and §4.4, Eq. (10)] The abstract's claim that the method estimates accurate global 3D poses 'from images only' conflicts with the assumptions stated in Section 4: known camera intrinsics, known relative camera positions, and limb lengths ℓ(i,j) that are taken from the ground truth data. Using test-subject ground-truth limb lengths in Eq. (10) constitutes evaluation leakage and makes the reported numbers more optimistic than a strict 'from images only' reading implies. The claims should be revised to state the actual required inputs, and the limb-length supervision should be either removed or explicitly justified as a manual measurement that would be available in practice.
  3. [§5.2, Table 2] The main comparison baseline, 'Ostrek et al. (unpublished data)', is not publicly available or citable. This makes the claim that the proposed method performs comparably to this baseline unverifiable. Please replace it with a published, accessible baseline or release the necessary implementation and experimental details so that the comparison can be reproduced.
  4. [§4.5, Eqs. (12)-(14)] The uncalibrated accuracy is critically dependent on the homography-derived rotation differences ΔR computed from ORB features in the static background. The introduction reports that off-the-shelf structure-from-motion fails on the same skiing footage 'due to the large zoom, large distance and view angle between cameras, and lack of discriminative patterns on the ski slope,' but the only validation of ΔR is a single qualitative plot (Fig. 5). Because the ablation Ours-B, which does not use Eq. (14), collapses to a global MPJPE of 7.59 m, the 0.701 m uncalibrated result is contingent on the background-feature conditioning. Please provide a quantitative evaluation of the ΔR estimates across cameras and sequences and a sensitivity analysis with respect to background texture, feature count, and moving objects, or explicitly restrict the claim to scenes with sufficiently textured static backgrounds.
minor comments (5)
  1. [Table 2] There is a typographical error in the Global speed MAE row: '1.87.±2.64' should presumably read '1.87±2.64'.
  2. [Table 1] In configuration (b), the MV-Ski all-joints MPJPE of 0.0137±0.1236 appears inconsistent with the corresponding PCK of 65.51; this is a much lower error than the SV-Ski all-joints value of 0.0807 despite a much lower PCK. Please verify these numbers.
  3. [§5, Fig. 5] Figure 5 shows estimated rotation differences for only one camera and one sequence; consider adding a summary of errors across all cameras and test sequences, or error bars.
  4. [§4.3, §4.5] The DCT basis sizes N_Π and N_Γ are chosen empirically (25/11 and 11), and the paper does not report how sensitive the results are to these values. A small sensitivity analysis would help the reader understand the robustness of the parametrization.
  5. [§4.4, Eq. (10)] Please clarify whether the limb lengths in Eq. (10) are measured from the same test subjects whose ground-truth poses are used for evaluation; if so, this should be explicitly stated as a limitation in the experiments section.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: camera rotations are measured from background features and pose is reconstructed by bundle adjustment, with no equation-level reduction to fitted inputs.

full rationale

The uncalibrated branch uses two independent image cues: 2D joint detections and per-frame relative camera rotations computed from ORB homographies on the static background (Eqs. 12-13). The rotation term E_rot (Eq. 14) enforces consistency between the DCT-parameterized rotations and these measured Delta-R values, while the reprojection term (Eq. 7) ties the DCT-parameterized pose to the 2D detections. These are measurement terms, not definitions of the quantities being predicted, and neither term is obtained from the other or from the 3D target. The limb-length prior uses ground-truth limb lengths as constants, which is an input prior rather than a fitted parameter renamed as a result, and the global position is not read off from it. The self-citations (Rhodin et al. [35] as a monocular baseline and for the NMPJPE definition; Sporri/Fasel [38,11] for skiing metrics) appear as comparisons and metric definitions, not as load-bearing support for the derivation. The Ours-B ablation shows that without E_rot the joint optimization collapses, making the homography measurement empirically load-bearing for accuracy; however, this is dependence on an independent sensor measurement, not circularity. The statement that off-the-shelf structure-from-motion fails on the ski footage is a robustness limitation, not a circular step. The inclusion of one MV-Ski sequence in 2D detector training is an evaluation-leakage risk, but the test set still uses a held-out skier and the 3D derivation itself remains self-contained: pose and camera rotation are jointly optimized against independent reprojection and background-motion observations.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The method rests on static-background homographies, known intrinsics and positions, smooth-motion DCT priors, and fixed limb lengths. No new physical entities are introduced. The free parameters are hyperparameters chosen on validation or by hand, and the paper does not provide a sensitivity analysis, so the contribution is an engineered integration rather than a parameter-free derivation.

free parameters (6)
  • DCT basis size for pose N_Pi = 25 (calibrated), 11 (uncalibrated)
    Chosen empirically in Section 4.3 as a trade-off between smoothness and ability to represent fast motions.
  • DCT basis size for camera rotation N_Gamma = 11
    Set to 11 in Section 4.5; no sensitivity analysis is provided.
  • Reprojection weight lambda_rep = 80 (calibrated), 500 (uncalibrated)
    Hand-tuned weights in Eqs. 6 and the uncalibrated variant in Section 4.5.
  • Limb length weight lambda_limbs = 1
    Set in Section 4.4.
  • Rotation consistency weight lambda_rot = 10000
    Set in Section 4.5; this strong weight is critical to the uncalibrated method's convergence.
  • Robust norm variance sigma^2 = 100
    Set in Section 4.4 to define the claimed outlier suppression.
assumptions (6)
  • domain assumption PTZ cameras rotate about a fixed optical center with no translation
    The homography model in Eq. 12 assumes inter-frame motion is a pure rotation; camera shake or translation would violate this. Section 4.5.
  • domain assumption The background outside the athlete's bounding box is static and sufficiently textured for ORB feature matching
    The rotation differences rely on features tracked in the background; low-texture slopes or moving spectators would break the homography estimate. Section 4.5.
  • domain assumption Camera intrinsics K and relative camera positions t are known
    The pipeline assumes intrinsically calibrated cameras with known relative position; only orientation is estimated. Section 4.
  • domain assumption Human limb lengths are constant and available (from ground truth or manual measurement)
    The prior in Eq. 10 fixes limb lengths; in the experiments they are taken from the dataset's ground truth. Section 4.4.
  • domain assumption Athlete motion and camera rotation are smooth and band-limited, so a low-order DCT basis represents them
    The IDCT parameterization (Eq. 1) suppresses high-frequency motion; very fast or jerky movements would be smoothed away. Sections 4.3 and 4.5.
  • domain assumption The 2D pose detector's heatmap maximum is a reliable joint location
    The method treats the argmax of each heatmap as the joint detection; systematic detector biases propagate into the 3D optimization. Section 4.2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Motion Capture from Pan-Tilt Cameras with Unknown Orientation." pith.science (2026). https://pith.science/paper/JX33PGTT

@misc{pith2026190811676,
  author       = {Pith},
  title        = {Pith review of: Motion Capture from Pan-Tilt Cameras with Unknown Orientation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JX33PGTT}},
  note         = {Machine review of arXiv:1908.11676}
}
read the original abstract

In sports, such as alpine skiing, coaches would like to know the speed and various biomechanical variables of their athletes and competitors. Existing methods use either body-worn sensors, which are cumbersome to setup, or manual image annotation, which is time consuming. We propose a method for estimating an athlete's global 3D position and articulated pose using multiple cameras. By contrast to classical markerless motion capture solutions, we allow cameras to rotate freely so that large capture volumes can be covered. In a first step, tight crops around the skier are predicted and fed to a 2D pose estimator network. The 3D pose is then reconstructed using a bundle adjustment method. Key to our solution is the rotation estimation of Pan-Tilt cameras in a joint optimization with the athlete pose and conditioning on relative background motion computed with feature tracking. Furthermore, we created a new alpine skiing dataset and annotated it with 2D pose labels, to overcome shortcomings of existing ones. Our method estimates accurate global 3D poses from images only and provides coaches with an automatic and fast tool for measuring and improving an athlete's performance.

Figures

Figures reproduced from arXiv: 1908.11676 by the authors.

Figure 1
Figure 1. Setup overview of the multi-view skiing dataset. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. Left: Example image showcasing calibration [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Method overview – Images from up to 6 different cameras are preprocessed to find 2D joint locations by first [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Estimated change in rotation ∆R for one camera. with Rc = IDCT(Γ c ) in DCT encoding, minimizing the norm between the measured ∆Rf,c and estimated consecu￾tive camera rotations Rf+1,c Rf,c| . By minimizing this term, we enforce that the estimated relative camera motio…
Figure 6
Figure 6. Figure 6: Comparison of all four dataset configurations, showing the improvements each addition yields on unseen test sets [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: With known rotation matrices: Comparison of all performance metrics for different number of cameras used in the [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: Estimating rotation matrices: Comparison of all performance metrics for different number of cameras used in the [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 43 canonical work pages

  1. [1]

    https: //docs.opencv.org/3.0-beta/modules/ calib3d/doc/camera_calibration_and_3d_ reconstruction.html

    Camera Calibration and 3D Reconstruction. https: //docs.opencv.org/3.0-beta/modules/ calib3d/doc/camera_calibration_and_3d_ reconstruction.html. [Online; last accessed on Jun 10, 2019]. 6

    work page 2019

  2. [2]

    https://docs.opencv.org/3

    Feature Matching. https://docs.opencv.org/3. 0-beta/doc/py_tutorials/py_feature2d/py_ matcher/py_matcher.html. [Online; last accessed on Jun 10, 2019]. 6

    work page 2019

  3. [3]

    https: //docs.opencv.org/3.0-beta/doc/py_ tutorials/py_feature2d/py_orb/py_orb

    ORB (Oriented FAST and Rotated BRIEF). https: //docs.opencv.org/3.0-beta/doc/py_ tutorials/py_feature2d/py_orb/py_orb. html. [Online; last accessed on Jun 10, 2019]. 6

    work page 2019

  4. [4]

    Akhter, T

    I. Akhter, T. Simon, S. Khan, I. Matthews, and Y . Sheikh. Bilinear spatiotemporal basis models. ACM Transactions on Graphics (TOG), 31(2):17, 2012. 5

    work page 2012

  5. [5]

    Andriluka, L

    M. Andriluka, L. Pishchulin, P. Gehler, and B. Schiele. 2D Human Pose Estimation: New Benchmark and State of the Art Analysis. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2014. 2

    work page 2014

  6. [6]

    Z. Cao, T. Simon, S. Wei, and Y . Sheikh. Realtime Multi- Person 2D Pose Estimation using Part Affinity Fields.CVPR,

    work page

  7. [7]

    Sports Camera Calibration via Synthetic Data

    J. Chen and J. J. Little. Sports Camera Calibration via Syn- thetic Data. CoRR, abs/1810.10658, 2018. 2

    work page Pith review arXiv 2018

  8. [8]

    J. Chen, F. Zhu, and J. J. Little. A Two-point Method for PTZ Camera Calibration in Sports. WACV, 2018. 2

    work page 2018

Show all 44 references
  1. [9]

    deGroot and E

    M. deGroot and E. Brown. SSD: Single Shot Multi- Box Object Detector, in PyTorch. https://github. com/amdegroot/ssd.pytorch. [Online; using com- mit 8dd3865 on Mar 30, 2018]. 16

    2018

  2. [10]

    Elhayek, C

    A. Elhayek, C. Stoll, K. I. Kim, and C. Theobalt. Outdoor Human Motion Capture by Simultaneous Optimization of Pose and Camera Parameters. Computer Graphics Forum, 34, 12 2014. 2

    2014

  3. [11]

    Fasel, J

    B. Fasel, J. Sp ¨orri, M. Gilgien, G. Boffi, J. Chardonnens, E. M ¨uller, and K. Aminian. Three-Dimensional Body and Centre of Mass Kinematics in Alpine Ski Racing Using Dif- ferential GNSS and Inertial Sensors. Remote Sensing, 8, 09

  4. [12]

    Gilgien, J

    M. Gilgien, J. Sp ¨orri, J. Chardonnes, J. Kr ¨oll, P. Limpach, and E. M ¨uller. Determination of the centre of mass kine- matics in alpine skiing using differential global navigation satellite systems. Journal of Sports Sciences , 33(9):960-9,

  5. [13]

    Gilgien, J

    M. Gilgien, J. Sp ¨orri, P. Limpach, A. Geiger, and E. M¨uller. The effect of different Global Navigation Satellite System methods on positioning accuracy in elite alpine skiing. Sen- sors (Basel), 14(10):18433-53, 2014. 1

    2014

  6. [14]

    G ¨unel, H

    S. G ¨unel, H. Rhodin, and P. Fua. What face and body shapes can tell about height. arXiv preprint arXiv:1805.10355 ,

    arXiv

  7. [15]

    Gwangjae, J

    Y . Gwangjae, J. J. Young, K. Jinhyeok, H. K. Jin, Y . K. Hye, K. Kitae, and B. P. Siddhartha. Potential of IMU Sensors in Performance Analysis of Professional Alpine Skiers. Sen- sors (Basel), 16(4):463, 2016. 1

    2016

  8. [16]

    Hartley and A

    R. Hartley and A. Zisserman. Multiple View Geometry in Computer Vision. Cambridge University Press, New York, NY , USA, 2 edition, 2003. 6

    2003

  9. [17]

    Hasler, B

    N. Hasler, B. Rosenhahn, T. Thormahlen, M. Wand, J. Gall, and H.-P. Seidel. Markerless motion capture with unsynchro- nized moving cameras. pages 224 – 231, 07 2009. 2

    2009

  10. [18]

    Huang, F

    C. Huang, F. Gao, J. Pan, Z. Yang, W. Qiu, P. Chen, X. Yang, S. Shen, and K. Cheng. Act: An autonomous drone cine- matography system for action scenes. 2018. 2

    2018

  11. [19]

    Insafutdinov, L

    E. Insafutdinov, L. Pishchulin, B. Andres, M. Andriluka, and B. Schiele. Deepercut: A Deeper, Stronger, and Faster Multi- Person Pose Estimation Model. 2016. 2

    2016

  12. [20]

    Ionescu, D

    C. Ionescu, D. Papava, V . Olaru, and C. Sminchisescu. Human3.6M: Large Scale Datasets and Predictive Methods for 3D Human Sensing in Natural Environments. IEEE Transactions on Pattern Analysis and Machine Intelligence, 36(7):1325–1339, jul 2014. 2

    2014

  13. [21]

    D. C. Liu and J. Nocedal. On the limited memory BFGS method for large scale optimization. Math. Program. , 45:503–528, 1989. 6

    1989

  14. [22]

    W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. E. Reed, C. Fu, and A. C. Berg. SSD: Single Shot MultiBox Detector. ECCV, 2016. 3, 4

    2016

  15. [23]

    Martinez, R

    J. Martinez, R. Hossain, J. Romero, and J. Little. A Simple Yet Effective Baseline for 3D Human Pose Estimation. 2017. 1

    2017

  16. [24]

    Mehta, H

    D. Mehta, H. Rhodin, D. Casas, P. Fua, O. Sotnychenko, W. Xu, and C. Theobalt. Monocular 3D Human Pose Esti- mation in the Wild Using Improved CNN Supervision. 2017. 1

    2017

  17. [25]

    Moeslund and E

    T. Moeslund and E. Granum. A Survey of Computer Vision- Based Human Motion Capture. 81(3), March 2001. 2

    2001

  18. [26]

    T. B. Moeslund, A. Hilton, and V . Kr¨uger. A Survey of Ad- vances in Vision-Based Human Motion Capture and Analy- sis. 104(2):90–126, 2006. 2

    2006

  19. [27]

    Moreno-noguer

    F. Moreno-noguer. 3D Human Pose Estimation from a Single Image via Distance Matrix Regression. 2017. 1

    2017

  20. [28]

    Nemec, T

    B. Nemec, T. Petri ˇc, J. Babi ˇc, and M. Supej. Estimation of alpine skier posture using machine learning techniques. Sensors (Basel), 14(10):18898-914, 2014. 1

    2014

  21. [29]

    Pavlakos, X

    G. Pavlakos, X. Zhou, K. Derpanis, G. Konstantinos, and K. Daniilidis. Coarse-To-Fine V olumetric Prediction for Single-Image 3D Human Pose. 2017. 1

    2017

  22. [30]

    Pavlakos, X

    G. Pavlakos, X. Zhou, K. D. G. Konstantinos, and D. Kostas. Harvesting Multiple Views for Marker-Less 3D Human Pose Annotations. 2017. 1, 2

    2017

  23. [31]

    A.-I. Popa, M. Zanfir, and C. Sminchisescu. Deep Multi- task Architecture for Integrated 2D and 3D Human Sensing

  24. [32]

    Puwein, L

    J. Puwein, L. Ballan, R. Ziegler, and M. Pollefeys. Joint Camera Pose Estimation and 3D Human Pose Estimation in a Multi-camera Setup. In ACCV, 2014. 2

    2014

  25. [33]

    Puwein, R

    J. Puwein, R. Ziegler, L. Ballan, and M. Pollefeys. PTZ Camera Network Calibration from Moving People in Sports Broadcasts. pages 25–32, 01 2012. 2

    2012

  26. [34]

    Rhodin, N

    H. Rhodin, N. Robertini, D. Casas, C. Richardt, H.-P. Sei- del, and C. Theobalt. General Automatic Human Shape and Motion Capture Using V olumetric Contour Cues. 2016. 2 9

    2016

  27. [35]

    Rhodin, J

    H. Rhodin, J. Sp ¨orri, I. Katircioglu, V . Constantin, F. Meyer, E. M ¨uller, M. Salzmann, and P. Fua. Learning Monocular 3D Human Pose Estimation from Multi-view Images.CVPR,

  28. [36]

    Rogez, P

    G. Rogez, P. Weinzaepfel, and C. Schmid. Lcr-Net: Localization-Classification-Regression for Human Pose

  29. [37]

    Rublee, V

    E. Rublee, V . Rabaud, K. Konolige, and G. Bradski. Orb: An efficient alternative to sift or surf. In Proceedings of the 2011 International Conference on Computer Vision , ICCV ’11, pages 2564–2571, Washington, DC, USA, 2011. IEEE Computer Society. 6

    2011

  30. [38]

    Sp ¨orri

    J. Sp ¨orri. Reasearch Dedicated to Sports Injury Prevention - the ’Sequence of Prevention’ on the Example of Alpine Ski Racing. Habilitation with Venia Docendi in Biomechanics,

  31. [39]

    X. Sun, J. Shang, S. Liang, and Y . Wei. Compositional Hu- man Pose Regression. 2017. 1

    2017

  32. [40]

    Takahashi, D

    K. Takahashi, D. Mikami, M. Isogawa, and H. Kimata. Hu- man Pose As Calibration Pattern; 3D Human Pose Estima- tion With Multiple Unsynchronized and Uncalibrated Cam- eras. In The IEEE Conference on Computer Vision and Pat- tern Recognition (CVPR) Workshops, June 2018. 2, 5, 6, 8

    2018

  33. [41]

    Tekin, P

    B. Tekin, P. Marquez-neila, M. Salzmann, and P. Fua. Learn- ing to Fuse 2D and 3D Image Cues for Monocular Body Pose Estimation. 2017. 1

    2017

  34. [42]

    D. Tome, C. Russell, and L. Agapito. Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image. In arXiv preprint, arXiv:1701.00295, 2017. 1

    2017 arXiv

  35. [43]

    Zanfir, E

    A. Zanfir, E. Marinoiu, and C. Sminchisescu. Monocular 3D Pose and Shape Estimation of Multiple People in Natural Scenes - the Importance of Multiple Scene Constraints. June

  36. [44]

    X. Zhou, Q. Huang, X. Sun, X. Xue, and Y . We. Weakly- Supervised Transfer for 3D Human Pose Estimation in the Wild. 2017. 1 10 Appendices In this supplementary material, we analyze the performance gains of using different dataset configurations in both 2D and 3D pose estimatio...

    2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.